KR102447691B1 - Determining Contributions of Various User Interactions to a Conversion - Google Patents

Abstract

The online system maintains information and other content describing user interactions with advertisements. Additionally, the online system maintains a confidence value each associated with the various types of information identifying the user and the information type identifying the user that provides a measure of the likelihood of the type of information identifying a particular user. Exemplary types of information identifying a user include an online system identifier, a client device identifier, and an application identifier. When the online system receives information indicating that the user has performed a conversion, the online system retrieves an interaction associated with at least one type of information identifying the user. An attribution model is applied to the retrieved interactions to determine the contribution of the various interactions to conversion, taking into account the confidence values associated with the types of information that identify the users associated with the various interactions.

Description

Determining Contributions of Various User Interactions to a Conversion

FIELD OF THE INVENTION The present invention relates generally to monitoring conversions, and more particularly to determining the contribution of interactions to conversions performed.

A content provider may associate a conversion with content provided by the content provider for presentation to an online system. A conversion identifies an interaction or type of interaction that the content provider wants to perform when users of the online system are presented with content. For example, a content provider associates content with a transition such that a user accesses a website related to the content presented by the online system or installs an application identified by the content presented by the online system.

However, users often perform a number of interactions captured by the online system prior to performing conversions related to the content presented by the online system. The various interactions performed by the user may increase or decrease the likelihood that the user will perform a conversion. However, the information that is communicated to the online system describing the different interactions the user has performed may include different types of information identifying the user. Due to the use of these different types of user identification information, conventional online systems do not accurately identify previous interactions by the user that may have contributed to the conversion performed by the user.

In order to provide advertisers with more accurate information about the contribution of various interactions by users to conversion, the online system identifies interactions by users prior to conversion and determines the contribution of different interactions to conversion. . The interaction's contribution to conversion provides a measure of the interaction's impact on the user performing the conversion. The online system stores information that describes the various interactions performed by the user, as well as information related to the conversions identified with the online system. Examples of information related to an interaction with an online system include: a description of the type of interaction (eg, exposure of an advertisement, interaction with an advertisement or link, a search performed online, an online purchase, etc.), the interaction information identifying the user who performed it (eg, online system user identifier, device identifier, etc.), and the time (eg, timestamp) associated with the interaction. The online system also associates confidence values with various types of user identification information that represent a measure of the information type's ability to accurately identify a user. For example, an online system associates a trust value with an online system user identifier that is higher than a trust value associated with the device identifier when multiple users share a device. In some embodiments, the online system ranks information types that identify users based on their confidence values, a higher position in the ranking indicates a higher ability of an information type to accurately identify a user. When the online system receives information describing the transition, the online system identifies information identifying the user from the received information. The online system then identifies interactions related to information identifying the user as well as interactions related to other types of information identifying the user.

An attribution model is applied to the retrieved interactions to evaluate the contribution of sources to conversion. In some embodiments, the attribution model weights different interactions to identify the contribution of various interactions to conversion. In various embodiments, the weight is based, at least in part, on a confidence value associated with a type of information identifying the user. Further, the attribution model may be applied to a subset of the retrieved interactions, such as the interactions retrieved with respect to a type of information identifying a user having at least one threshold confidence value.

Included in the context of the present invention.

1 is a block diagram of a system environment in which an online system is driven according to an exemplary embodiment.
2 is a block diagram of an online system according to an embodiment.
3 is a flow diagram of a method of contributing to a conversion among one or more interactions in accordance with one embodiment.
4 is an example of information associating interactions and conversions with types of information identifying an acting user, according to one embodiment.
The drawings depict various embodiments for illustrative purposes only. Those skilled in the art will readily recognize from the following description that alternative embodiments of the structures and methods shown herein may be utilized without departing from the principles disclosed herein.

system structure

1 is a block diagram of a system environment 100 for an online system 140 . The system environment 100 illustrated in FIG. 1 includes one or more client devices 110 , a network 120 , one or more third party systems 130 , and an online system 140 . In alternative configurations, different and/or additional components may be included in the system environment 100 .

Client device 110 is one or more computing devices capable of receiving user input and transmitting and/or receiving data over network 120 . In one embodiment, the client device 110 is a conventional computer system, such as a desktop or laptop computer. Alternatively, the client device 110 may be a device having computing capabilities, such as a personal digital assistant (PDA), mobile phone, smart phone, or other suitable device. The client device 110 is configured to communicate over the network 120 . In one embodiment, the client device 110 runs an application that allows a user of the client device 110 to interact with the online system 140 . For example, the client device 100 runs a browser application that enables interaction between the client device 100 and the online system 140 over the network 120 . In yet another embodiment, the client device 110 interacts with the online system 140 via an application programming interface (API) executing on the native operating system of the client device 110 , such as iOS® and ANDROID ^™ . do. A client device 110 may be associated with one or more users. In some embodiments, a user may access multiple applications, a third party system 130 , or any combination thereof through a client device 110 associated with the user.

Client device 110 is configured to communicate over network 120 , which may include any combination of local area networks and/or wide area networks, using both wired and wireless communication systems. In one embodiment, network 120 uses standard communication technologies and/or protocols. For example, network 120 includes communication links using technologies such as Ethernet, 802.11, worldwide interoperability for microwave access (WiMAX), 3G, 4G, code division multiple access (CDMA), digital subscriber line (DSL), etc. . Examples of networking protocols used for communication over network 120 include multiprotocol label switching (MPLS), transmission control protocol/Internet protocol (TCP/IP), hypertext transport protocol (HTTP), simple mail transfer protocol (SMTP), and Includes file transfer protocol (FTP). Data exchanged in network 120 may be represented using any suitable format, such as Hypertext Markup Language (HTML) or Extensible Markup Language (XML). In some embodiments, all or part of the communication link of network 120 may be encrypted using any suitable technology(s).

One or more third party systems 130 may be coupled to a network 120 for communicating with an online system 140 , which is further described below in conjunction with FIG. 2 . In one embodiment, the third-party system 130 communicates information describing an application to be executed on the client device 110 or communicates data to the client device 110 for use in an application running on the client device. application provider. In another embodiment, the third party system 130 provides content or other information for display via the client device 110 . In addition, the third-party system 130 may transmit information, such as advertisements, contents, or information about an application provided by the third-party system 130 to the online system 140 .

2 is a block diagram of the structure of the online system 120 . The online system 140 shown in FIG. 2 includes an action log 210a , a filtering module 215 , a contribution module 220 , and a web server 225 . In other embodiments, the online system 140 may include additional modules, fewer modules, or other modules for various applications. Conventional components such as network interfaces, security functions, load balancers, failover servers, management and network operation consoles, etc. are not shown in order not to obscure the details of the system architecture.

Each user of the online system 120 is associated with a user profile that is stored in a user profile store 205 . The user profile includes declarative information about the user that has been explicitly shared by the user, and may also include profile information inferred by the online system 140 . In one embodiment, a user profile includes a plurality of data fields each describing one or more attributes of a corresponding online system user. Examples of information stored in a user profile include, for example, biographical information such as career, education, gender, hobbies or preferences, location, etc., demographic information, and other types of descriptive information. In addition, the user profile in the user profile store 205 may manage a reference to the action stored in the action log 210 .

In various embodiments, multiple physical locations associated with the user are included in the user's user profile. A location type is associated with each physical location and describes one or more characteristics of the physical location. For example, the location type specifies the source from which the physical location was determined. In one embodiment, the location type associated with the physical location is selected from a set of location types. For example, a location type identifies a fixed physical location (eg, hometown, residence) that is specified by the user or inferred from information related to the user, and another location type is provided by the user to the online system 140 Identifies a physical location identified in the content (eg, a location identified as a post by the user to the online system 140 , a location the user checked in via the client device 110 ), wherein the additional location type is associated with the user The physical location identified by the information passed by the client device 110 to the online system 140 (eg, delivered to the online system 140 by the client device 110 running an application associated with the online system 140 ) GPS (Global Positioning System) coordinates) are identified. However, in other embodiments, the location type associated with the physical location may identify any suitable combination of properties of the physical location. Also, in some embodiments, the location type identifies the user's current physical location. For example, the online system 140 may include information communicated to the online system 140 by a client device 110 associated with the user and/or information provided to the online system 140 by the user (eg, as described above). It is determined whether the physical location associated with the user is the current physical location of the user by comparing the timestamp associated with the information received from the client device 110 to the current time (via content). If the difference between the times compared is less than the threshold, the online system 140 identifies the physical location associated with the user as the user's current location.

By associating different location types with different physical locations, a user profile can provide an indication of how likely a user is currently associated with different physical locations. For example, a physical location associated with a location type indicating that the physical location has been identified with information received by the online system 140 from the client device 110 may be associated with a location type indicating that the physical location is a user-specific fixed physical location and a location type indicating that the physical location is a user-specific fixed physical location. It is relatively more likely to be associated with the user at the current time than the associated physical location. The online system 140 may determine a location type associated with the physical location based on additional information received with the physical location. In some embodiments, metadata specifying the location type is received by the online system 140 along with the physical location.

The action log 210 may store user actions that are performed on a third-party system 130 , such as an external website, and communicated to the online system 140 . For example, an e-commerce website may recognize a user of the online system 140 via a social plug-in that enables the e-commerce website to identify a user of the online system 140 . Because users of the online system 140 are uniquely identifiable, as in the example above, the e-commerce website may use the online system ( 140) can be communicated with. Accordingly, the behavior log 210 records webpage browsing history, associated advertisements, purchases made, subscriptions to third-party services (eg, movie subscriptions, music subscriptions, e-book subscriptions, etc.) and other patterns from shopping and purchases. Including, information on actions performed by the user in the third-party system 130 may be recorded.

Behavior log 210 contains information describing interactions with content by online system users as well as conversions related to online system users. A conversion is an action requested by the advertiser and performed by the user, and may result from one or more interactions of the user specified by the advertiser to be monitored or tracked. In one embodiment, the behavior log 210 includes various databases that store information related to interactions and conversions. For example, the behavior log 210 includes a write-optimized database that stores information related to a number of interactions and conversions associated with the advertiser. As a further example, the behavior log 210 includes a read-optimized database that stores information related to a number of interactions related to user identifying information corresponding to a user identifying information related to conversions. Alternatively, the behavior log 210 includes a single database optimized for the most frequent types of interactions; For example, the behavior log 210 includes a write-optimized database in implementations where writing to the database is the most common behavior. By including different types of databases, the behavior log 210 can optimize the search time for any type of information and optimize the storage of other types of information.

Examples of types of information related to interactions or conversions stored in the action log 210 include: a description of the type of interaction (eg, exposure of an advertisement, access to an advertisement or link, searches performed by the user, previous conversions associated with the user) ) and the time (eg, timestamp) associated with the interaction or conversion. In addition, the behavior log identifies the user who performed the interaction or conversion by associating the user identification information with each interaction or conversion. Many types of information identifying a user may be associated with an interaction or conversion. Exemplary types of information identifying a user include an online system user identifier, a client device identifier, an application identifier, and any other suitable information.

The filtering module 215 indexes information related to interactions and conversions by advertisers into a database. For example, the filtering module 215 retrieves information related to multiple interactions and conversions associated with various advertisers from the behavior log 210 and indexes the retrieved information into a database associated with each advertiser. In addition, the filtering module 215 filters the interaction related to the conversion from the action log 210 based on the user identification information related to the conversion and the user identification information related to the interaction. For example, the filtering module 215 retrieves from the action log 210 information about interactions related to user identification information that matches user identification information related to conversions, and stores the retrieved information in a database for later retrieval. .

When retrieving information about interactions associated with user identification information related to conversion from the behavior log 210 , the filtering module 215 may retrieve information related to various types of information identifying the user from the behavior log 210 . can For example, online system 140 associates users with various types of user identification information (eg, online system user identifiers, device identifiers, application identifiers). If the type of user identification information associated with the conversion matches the user identification information associated with the conversion, the filtering module 215 may also determine an additional type of information identifying the user. Interactions related to additional types of user identification information are retrieved by the filtering module 215 , which enables retrieval of interactions related to multiple types of user identification information about a user.

The attribution module 220 includes one or more attribution rules applied to the identified conversion to identify the contribution of other interactions to the conversion. In one embodiment, the filtering module 215 passes the user identification information associated with the conversion and the interaction associated with one or more interaction types associated with at least one type of information identifying the user to the attribution module 220 . The attribution module 220 includes confidence values associated with different types of user identification information, wherein the confidence values associated with one type of user identification information are one type of user identification information that identifies a particular user of the online system 140 . provides a measure of the likelihood of For example, an online system user identifier is associated with a greater confidence value than a client device identifier because the online system user identifier is more likely to identify a particular user of the online system 140 . Associating a confidence value with a type of user interaction is further described below in conjunction with FIG. 3 .

In some embodiments, the attribution module 220 associates a weight with the various interactions received from the filtering module 215 based on types of user identification information associated with the different interactions. For example, a weight associated with an interaction is proportional to a confidence value associated with the type of information identifying the user associated with the interaction, so that an interaction associated with a type of user identification information with a higher confidence value is associated with a higher weight. related The attribution module 220 takes into account the weights associated with the various interactions when determining the contribution of the various interactions to a conversion. In some embodiments, the attribution model 220 associates weights with various interactions received from the filtering module 215 based on users and/or other information related to the interactions. For example, the attribution module 220 may employ a “time decay” model that reduces the importance of ad impressions, clicks, or other interactions based on the amount of time that has elapsed relative to the time when the conversation occurred. include

Web server 225 links online system 140 via network 120 to one or more client devices 110 as well as one or more third-party systems 130 . The web server 225 serves web pages as well as other content such as JAVA®, FLASH®, XML, and the like. The web server 225 may provide an online system 140 , such as an instant message, a queued message (eg, e-mail), a text message, an SMS (Short Message Service) message, or a message sent using any other suitable messaging technology. ) and the client device 110 may receive and route messages. A user may upload information (eg, an image or video) stored in the content store 210 by sending a request to the web server 225 . Additionally, web server 225 may provide application programming interface (API) functionality to send data directly to a native client device operating system, such as, for example, IOS®, ANDROID™, WEBOS® or Blackberry OS.

Contribution of interaction to conversion decision

3 is a flow diagram of one embodiment of a method for determining the contribution of various interactions to a conversion associated with an advertisement. In other embodiments, the method includes different, more, or fewer steps than shown in FIG. 3 . Additionally, in some embodiments, the steps described with respect to FIG. 3 may be performed in a different order.

The online system 140 stores 305 confidence values associated with various types of information that identify users of the online system. A confidence value associated with a type of information that identifies a user provides a measure of the likelihood that that information type will correctly identify the user. Examples of types of information identifying a user include: an online system user identifier, a client device identifier, and an application identifier. In one embodiment, the online system 140 ranks the types of information identifying users based on their associated confidence values. For example, a ranking of a type of user identification information has a user identification information type that is more likely to correctly identify a user at a higher position in the ranking (ie, a user identification information type associated with a higher confidence value).

The online system 140 also receives 310 information describing one or more interactions associated with a user of the online system 140 and stores 315 information describing the one or more interactions. For example, the online system 140 receives 310 information indicating that a user of the online system 140 has been presented with an advertisement, via a tracking pixel 310 embedded in the advertisement, 310 , an online system user identifier, a browser identifier. and a user having a client device identifier and indicating a time at which the user was presented with an advertisement. Information describing user interactions may be received in real-time or near real-time. In some embodiments, the online system 140 receives 310 and stores 315 information describing interactions with entities other than users of the online system 140 . For example, the online system 140 receives 310 information identifying an interaction with an advertisement, a timestamp associated with the interaction, and a client device identifier associated with the client device 110 used to perform the interaction. . In some embodiments, the online system 140 stores information describing one or more interactions with advertisements in a write-optimized database to more efficiently store information describing the interactions.

The online system 140 receives ( 320 ) information related to an advertisement and describing a conversion related to the user in action. As noted above, a conversion is a type of action associated with an advertisement by an advertiser as the type of interaction desired in response to presentation of the advertisement. Examples of conversions include purchases, interactions with advertisements, or other suitable interactions involving the user and advertisements in action of the online system 140 . Information describing the transition may be received 320 from a tracking pixel or other tracking mechanism of the third party system 130 or from any suitable source. Additionally, information describing the transition may be received in real-time or near real-time. The information describing the transition includes information identifying the acting user (eg, online system user identifier, client device identifier, application identifier), a description of the transition and the time associated with the transition. The online system 140 stores 325 information describing the transition. For example, the online system 140 stores information describing conversions and other interactions with the advertiser related to the conversion in a write-optimized database of the action log 210 .

Information identifying the acting user is identified 330 from the information describing the conversion, and information describing one or more additional interactions associated with the acting user is retrieved 335 using the information identifying the acting user. . Also, in some embodiments, if the information identifying the acting user is a specific type of information identifying the user, another type of information identifying the acting user is identified, and another type of information identifying the acting user is Interactions related to the information are also retrieved 335 . For example, if the information identifying the acting user is a device identifier, then the online system 140 identifies another type of information identifying the acting user (eg, online system user identifier, application identifier) and a trade name associated with the device identifier. Retrieve 335 interactions associated with other types of information that identify the user acting with the action.

In one embodiment, the information received by the online system 140 and describing the conversion and one or more interactions includes the type of interaction performed, one or more types of information identifying the user performing the interaction or conversion, and each It is stored in a write-optimized database that contains information about the times associated with interactions or transitions. When the online system 140 receives 320 information describing the conversion, one or more types of user identification information associated with the user acting are identified 330 from the received information and included in the information describing the conversion. is used to retrieve 335 one or more additional interactions related to the acting user having at least one information type identifying the acting user that matches the type of information identifying the user. In some embodiments, information describing one or more additional interactions retrieved is stored in a read-optimized database that expedites subsequent retrieval of information by the online system 140 .

In some embodiments, information identifying the user associated with the interaction and/or conversion may be inferred. If at least a threshold number or percentage of interactions associated with the type of information identifying the user also relate to the type of additional information that identifies the user, then the interaction with the type of information that identifies the user is also associated with the type of additional information that identifies the user. It is also inferred to be related. For example, if at least a threshold percentage of interactions related to a device identifier are also related to an online system user identifier, then it is inferred that interactions related only to the device identifier are also related to the online system user identifier. Contributions to conversion of an interaction having an inferred relevance to a type of user identification information may be discounted.

In some embodiments, the online system 140 stores 340 the retrieved information describing the interaction in a location that may optimize further retrieval of the information. For example, the retrieved information is stored 340 in a separate database than the action log 210 to expedite subsequent access to the retrieved information. In some embodiments, the retrieved information is stored 340 in a read-optimized database that allows the online system 140 to more quickly access the retrieved information.

Based on the retrieved information describing the interactions, the online system 140 passes the interactions to an attribution model that determines the contribution values of the various interactions to conversions. In one embodiment, the online system 140 determines 345 the contribution of various interactions to the conversion. The weights relate to various interactions from the retrieved information. In one embodiment, a weight is associated with each interaction described by the retrieved information. Interaction-related weights provide a measure of the interaction's amount of contribution to conversion. In various embodiments, the weight associated with the interaction is based, at least in part, on the type of information identifying the user associated with the interaction. The weight associated with the interaction is related to the type of information identifying the user taking action associated with the interaction. It can be proportional to the reliability value. For example, an interaction associated with a type of user identification information having a higher confidence value is associated with a higher weight. In some embodiments, the online system 140 determines weights associated with the various interactions and adds the weights and interactions to an attribution model that determines 345 a contribution value of the various interactions to a conversion based at least in part on the weights. It conveys information that describes the action. In some embodiments, the online system weights the various interactions based on the user and/or other information related to the interaction (eg, the length of time elapsed since the interaction and the time at which the conversion occurred).

An attribution model, or one or more attribution rules, is applied to interactions corresponding to the retrieved information to determine 345 the contribution of various interactions to conversions. Attribution rules or attribution models may be applied to weights associated with the various interactions in determining 345 the contribution of the various interactions to a conversion. For example, interactions are ranked based on a confidence value regarding the type of identifying information of the user in action associated with the various interactions, and interactions having at least one critical position in the ranking can be attributed to the interaction's contribution to conversion. It is passed on to the attribution model that determines As another example, an interaction related to a type of information identifying an acting user having at least one threshold confidence value is passed to the attribution model. In various embodiments, an interaction's contribution to a conversion is based, at least in part, on a weight associated with the interaction. The online system 140 may generate an attribution report for an advertiser based on application of an attribution model or attribution rule that identifies the contribution of various interactions to a conversion.

Examples of Identification of Interactions

4 shows an example of information that associates an interaction with a type 405 of information identifying an acting user. In the example of FIG. 4 , the online system user identifier, client device identifier, and application identifier are types 405 of information identifying the acting user. For illustrative purposes, FIG. 4 illustrates an example in which online system 140 receives information describing a conversion 410 associated with an acting user including an online system user identifier and an application identifier associated with the acting user.

Based on the online system user identifier and application identifier included in the information describing the conversion 410 , the online system 140 retrieves interactions related to various types of information identifying the acting user. In the example of FIG. 4 , the online system 140 interacts with interactions 415 , 420 , 425 that are respectively associated with the acting user but are associated with a different type of information 405 identifying the acting user. ))). Interaction 415 relates to a client device identifier associated with the acting user. Interaction 420 relates to an online system user identifier, a client device identifier, and an application identifier, respectively, associated with the acting user. Interaction 425 also relates to an application identifier and a client device identifier associated with the acting user.

From the interactions related to the type 405 of information related to the user acting, the online system 140 identifies one or more interactions related to information identifying the user also related to the conversion 410 . In the example of FIG. 4 , interaction 420 is associated with an online system user identifier associated with a conversion, and interaction 425 is associated with an application identifier associated with conversion 410 . Accordingly, the online system 140 applies an attribution model or one or more attribution rules to the interaction 420 and interaction 425 to determine the contribution of both the interaction 420 and the interaction 425 to the conversion 410 . apply to all As discussed above in conjunction with FIG. 3 , weights may be associated with interactions 420 and 425 based on a confidence value associated with a type of information identifying the acting user associated with each interaction. For example, if the online system user identifier has a higher trust value than the application identifier, a greater weight is associated with interaction 420 rather than interaction 425 . In various embodiments, weights associated with interactions are used to determine the contribution of each interaction to conversion 410 by an attribution model or attribution rule.

summary

The foregoing description of embodiments of the present invention is provided for purposes of explanation, and is not intended to be exhaustive or to limit the patent rights to the precise forms disclosed. Those of ordinary skill in the art to which the present invention pertains will recognize that various modifications and variations are possible from the above disclosure.

Portions of this specification describe embodiments as algorithms and symbolic representations of operations on information. These algorithmic descriptions and expressions are commonly used by those of ordinary skill in the data processing field to effectively convey the substance of their work to others skilled in the art. will become These operations, which are described functionally, computationally or logically, are understood to be implemented by a computer program or equivalent electrical circuit, microcode, or the like. Also, it is often found that such arrangements of operations can be referred to as modules without loss of generality. The described operations and related modules may be implemented in software, firmware, hardware, or any combination thereof.

Any steps, operations, or processes described herein may be performed or implemented by one or more hardware or software modules or in combination with other devices. In one embodiment, a software module is implemented as a computer program product comprising a computer readable medium comprising computer program code executable by a computer processor to perform some or all of the steps, operations or processes described. do.

Further, embodiments of the present invention may relate to an apparatus for performing the operations herein. The apparatus may include a general purpose computing apparatus that may be specifically constructed for the requested purpose and/or selectively activated or reconfigured by a computer program stored on a computer. Such a computer program may be stored in a non-transitory tangible computer readable storage medium or any type of medium suitable for storing electronic instructions capable of being coupled to a computer system bus. Moreover, any computing systems referred to herein may include a single processor, or may be an architecture employing a multi-processor design for increased computing power.

Further, embodiments of the present invention may relate to products manufactured with the computing processes described herein. Such products may include information generated as a result of a computing process, wherein the information is stored in a non-transitory tangible computer-readable storage medium, any implementation of a computer program product or other data combination described herein. Examples may be included.

Finally, the language used herein has been principally chosen for readability and elucidation purposes, and may not be chosen to detail or limit patent rights. Accordingly, it is intended that the scope of the patents be limited not by this detailed description, but by any claims that this application files based thereon. Accordingly, the disclosure of the embodiments is intended to be illustrative, but not limiting, of the scope of the patents set forth in the following claims.

Claims (20)

storing, in the online system, a reliability value each associated with a plurality of identifier types that identify users of the online system;
receiving information that describes a conversion associated with an advertisement and includes an identifier identifying an active user of an online system associated with the conversion;
retrieving one or more interactions that may contribute to the conversion in association with each type of identifier that identifies an acting user of the online system associated with the conversion;
associating each of the retrieved one or more interactions with a weight; and
passing, based on the weight, the weight and the retrieved one or more interactions to an attribution model configured to assign a contribution amount to one or more of the retrieved one or more interactions for the conversion;
the confidence value associated with the identifier type provides a measure of the likelihood that the identifier type identifies a user of the online system, the plurality of identifier types comprising at least two of an online system user identifier, a client device identifier, and an application identifier;
wherein the weight associated with the retrieved interaction is based, at least in part, on a confidence value associated with a type of identifier that identifies an acting user of the online system associated with the conversion. The method of claim 1,
In the online system, storing a trust value each associated with a plurality of identifier types identifying users of the online system includes:
and ranking a plurality of identifier types identifying users of the online system based at least in part on a confidence value. The method of claim 1,
wherein the one or more interactions that may contribute to the conversion include one or more prior conversions associated with the user in action. The method of claim 1,
One or more interactions that may contribute to a conversion include: viewing one or more advertisements, interacting with one or more advertisements, accessing one or more objects associated with one or more advertisements, making one or more purchases associated with the advertiser for an advertisement. , requesting one or more services associated with the advertiser for advertising, participating in one or more events associated with the advertiser for advertising, subscribing to one or more third-party services, and any combination thereof. A method performed by a computer. The method of claim 1,
and determining a contribution amount of one or more of the retrieved one or more active user interactions for the conversion based at least in part on an attribution model and a weight and an attribution model each associated with the retrieved one or more interactions. 6. The method of claim 5,
Determining an amount of contribution of one or more of the retrieved one or more active user interactions to the conversion based at least in part on a weight and attribution model each associated with the retrieved one or more interactions, comprising:
At least in part based on a confidence value associated with a type of identifier that identifies an acting user associated with each of the one or more interactions retrieved, at least in part to contribute to a conversion with respect to each type of identifier that identifies an acting user of the online system associated with the conversion. identifying a subset of the retrieved one or more interactions that are possible; and
determining a contribution amount of each interaction in a subset of the one or more interactions retrieved for conversion based at least in part on the weight. 7. The method of claim 6,
Based at least in part on the credit value, identifying a subset of the retrieved one or more interactions that may contribute to a conversion with respect to each type of identifier that identifies an acting user of the online system associated with the conversion, comprising:
ranking the identifier type based at least in part on the confidence value; and
and selecting a retrieved interaction associated with an identifier type having at least one threshold position within the ranking. The method of claim 1,
Retrieving one or more interactions that may contribute to a conversion with respect to each type of identifier that identifies an acting user of the online system associated with the conversion comprises:
and retrieving an interaction associated with a plurality of identifier types that identify an acting user of the online system associated with the conversion. The method of claim 1,
Retrieving one or more interactions that may contribute to a conversion with respect to each type of identifier that identifies an acting user of the online system associated with the conversion comprises:
a plurality of identifier types that identify an acting user associated with one or more interactions of the acting user, the plurality of identifier types being associated with at least one threshold percentage of the one or more interactions of the acting user; identifying;
identifying one or more additional interactions of the acting user of the online system associated with a subset of the plurality of identifier types that identify the acting user;
associating one or more further interactions of the user with a plurality of identifier types identifying the user; and
retrieving one or more interactions that may contribute to a conversion with respect to at least one identifier type that identifies the user in action. receiving at the online system information describing a conversion associated with the advertisement and comprising information identifying a user of the online system associated with the conversion;
identifying, by the online system, one or more identifier types identifying the user based on the received information identifying the user, the one or more identifier types including an online system user identifier, a client device identifier, and an application identifier; Identifying one or more identifier types identifying the user based on the received information identifying the user comprises:
retrieving a confidence value each associated with one or more identifier types identifying a user of the online system, the confidence value associated with the identifier type providing a measure of the likelihood that the identifier type identifies a user of the online system;
searching by the online system for one or more interactions that may contribute to a conversion based at least in part on the one or more identified identifier types identifying the user; and
identifying, comprising communicating, by the online system, the retrieved one or more interactions to an attribution model configured to assign an amount of contribution to one or more of the retrieved one or more interactions for a conversion; and
determining, by the online system, a contribution amount of each of the one or more interactions that may contribute to a conversion, based at least in part on an attribution model and one or more identifier types identifying a user each associated with the one or more interactions, comprising: Determining, based at least in part on an attribution model and at least one identifier type identifying a user each associated with the one or more interactions, an amount of contribution of each of the one or more interactions that may contribute to a conversion comprises:
and associating a weight with one or more interactions based at least in part on a confidence value. 11. The method of claim 10,
Identifying one or more identifier types identifying the user based on the received information identifying the user comprises:
and ranking one or more identifier types identifying users of the online system based at least in part on the confidence value. 11. The method of claim 10,
wherein the retrieved one or more interactions that may contribute to the conversion include one or more previous conversions associated with the user. 11. The method of claim 10,
The retrieved one or more interactions that may contribute to a conversion include: viewing one or more advertisements, interacting with one or more advertisements, accessing one or more objects associated with one or more advertisements, making one or more purchases associated with the advertiser for an advertisement. in the group consisting of: requesting one or more services associated with the advertiser for advertising, participating in one or more events associated with the advertiser for advertising, subscribing to one or more third-party services, and any combination thereof; chosen way. 11. The method of claim 10,
Determining, based at least in part on the attribution model and one or more identifier types identifying users each associated with the one or more interactions, a contribution amount of each of the one or more interactions that may contribute to a conversion comprises:
determining a contribution of each of the user's one or more interactions to the conversion based at least in part on a weight associated with the one or more identifier types that identify the user associated with the one or more interactions that may contribute to the conversion; Including method. 11. The method of claim 10,
Determining, based at least in part on the attribution model and one or more identifier types identifying users each associated with the one or more interactions, a contribution amount of each of the one or more interactions that may contribute to a conversion comprises:
identifying, based at least in part on the confidence value, a subset of the retrieved one or more interactions that may contribute to the conversion; and
determining, based at least in part on the confidence value, a contribution of each interaction in a subset of the retrieved one or more interactions that may contribute to the conversion. 16. The method of claim 15,
Identifying, based at least in part on the confidence value, a subset of the retrieved one or more interactions that may contribute to the conversion comprises:
ranking the identifier type based at least in part on the confidence value; and
and selecting a retrieved interaction associated with an identifier type having at least one threshold position within the ranking. A computer-readable storage medium having recorded thereon encoded instructions, when executed by a processor, the instructions cause the processor to:
receiving information describing a conversion associated with the advertisement and comprising information identifying a user of an online system associated with the conversion;
Identifies one or more identifier types identifying a user based on received information identifying the user, the one or more identifier types including an online system user identifier, a client device identifier, and an application identifier, the receiving identifier identifying the user Identifying one or more types of identifiers identifying the user based on the identified information includes:
retrieving a confidence value each associated with one or more identifier types identifying a user of the online system, the confidence value associated with the identifier type providing a measure of the likelihood that the identifier type identifies a user of the online system;
retrieving one or more interactions that may contribute to a conversion based at least in part on the one or more identified identifier types identifying the user; and
identifying the retrieved one or more interactions, comprising passing the retrieved one or more interactions to an attribution model configured to assign an attribution amount to one or more of the retrieved one or more interactions for a conversion; and
determining, based at least in part on an attribution model and at least one identifier type identifying a user each associated with the one or more interactions, an attribution amount of each of the one or more interactions that may contribute to a conversion, wherein each of the one or more interactions and each Determining, based at least in part on an attribution model and one or more identifier types identifying the associated user, an amount of contribution of each of the one or more interactions that may contribute to a conversion comprises:
and associating a weight with one or more interactions based at least in part on a confidence value. 18. The method of claim 17,
Identifying one or more identifier types identifying the user based on the received information identifying the user comprises:
and ranking one or more identifier types that identify a user of the online system based at least in part on a confidence value.
delete delete

Images